The invention relates to a system and method for measuring periodic optical radiation signals by means of autocorrelation and more particularly to an improved system and method for autocorrelation of optical radiation signals using no moving parts.
In order to produce precise, repeatable pulses of laser radiation having very short durations of less than several picoseconds, it is necessary to accurately measure the laser output, most preferably by analyzing the shape of the pulses. Present laser technology permits the generation of waveforms consisting of pulses in the picosecond range, but such pulses are too short to be measured directly by conventional photodetectors. In order to obtain an accurate picture of such pulses it is necessary to employ a technique known as autocorrelation.
Autocorrelation can briefly be described as a measurement system for developing a composite image of a periodic waveform through multiplication of the measured signal with itself, over a range of time adjustments between the signals multiplied. The measured signal is first divided into two signals, and a time adjustment is introduced between the two signals by delaying one or the other or both signals by selected amounts. The signals are then multiplied together, and the magnitude of the product provides information about the original signal being measured. For laser signals, a beam splitter is used to divide the measured signal into two beams and then one or the other or both beams are directed through a prism or other medium, introducing a slight time adjustment between the beams. The beams are then recombined and a product value is measured. How the product value is used can best be illustrated by describing the measurement of a laser pulse train.
The individual pulses in a pulse train can be thought of as parts of a signal which is zero everywhere except over a very small interval, called the pulse width. When such a signal is multiplied with an identical pulse train which is precisely in phase with the first signal, the product will be another pulse train of essentially equal pulses. If, however, the two pulse trains are out of phase by more than their pulse widths within one, the product signal will be zero. When the pulse trains are out of phase by a fraction of their pulse widths, the product signal will be greater than zero, but less than when the signals exactly coincide. The object of autocorrelation is to produce a series of selected phase mismatches between two pulse beams which are created from the measured signal. At each selected phase mismatch the product is measured. A series of such measurements provides an image of the pulses in the pulse beam, which can then be displayed on an oscilloscope.
Prior art systems for autocorrelating laser and other optical radiation signals make use of varying thicknesses of optical material to produce the relative time adjustments between the optical beams. Optical material delays a beam which is passed through it by an amount dependent on its thickness and index of refraction. To vary the thickness, in most prior art autocorrelators, a pair of beams are directed through a rotating rectangular prism formed of optical material. When the prism is in one position, a short beam path is provided for one beam and a longer path for the other beam. In another position the prism will provide equal length paths through the prism for both beams. Thus, as the prism rotates the beams are delayed by different amount, relative to one another, over a range of such delays. After emerging from the prism, the beams are focused by conventional means on a single point where a product is derived.
While such prior art systems can effectively produce the requisite range of time adjustments between the beams necessary for autocorrelation, they do suffer from certain disadvantages. Because prior art autocorrelators employ a moving prism, a certain amount of mechanical motion and vibration is unavoidable. Even a precisely balanced prism rotating on a single motor shaft will generate mechanical disturbances unacceptably large for use within laser optical cavities. The vibrations produced by the rotating prism make it impossible to mount such an autocorrelator directly on a laser resonator structure. Instead, prior art autocorrelators must be mechanically isolated from the source of the signal being measured. Such isolation requires separate components to be precisely aligned, increasing set-up time and inconvenience.
It would be advantageous to be able to measure to a high degree of accuracy, the pulses generated in a pulse laser without the need for setup and alignment of separate components. More particularly, it would be desirable to be able to incorporate an autocorrelator into a laser resonator structure. Such an autocorrelator must include no moving parts which could generate unacceptable vibrations. Alternatively, a vibration-free autocorrelator could be detachably mounted on a laser resonator, facititating alignment with the laser.